The discussion of Junicode Wynn raises an interesting question: how can I find whether my system has a particular unicode character on it? For those who don't know, unicode is a universal coding system to describe characters (see http://www.unicode.org/). Before unicode, characters were largely associated with letters on the keyboard (I'm simplifying a bit). A predecessor of Peter Baker's Unicode font, Times Old English, used to work this way, for example. When you switched to that font in a wordprocessor and typed a letter (let's say "t"), some other character would appear on your screen (let's say thorn). The trouble was if you didn't have the right font, there was no way of recovering the missing character. What you intended as "thorn" would appear to your reader as "t".

Unicode solves this problem by assigning individual characters a unique code point (there are millions of these code points available in the system). This means in theory (and largely in practice) that you no longer need to rely on a user having exactly your font in order to see your character; in fact you don't even need to know which font they will use: as long as they have a font that has the character in it, your document will be readable.

In practice it isn't quite that simple, since many characters are not found in all or even many fonts. Wynn (01BF) is one of them. On my Windows XP system (which has hundreds of fonts installed), Wynn is present in only two: Junicode and Lucinda Sans Unicode. So while I can indicate that I mean wynn in a document by using the unicode code point 01BF, there is a good chance my user may not be able to see an actual wynn in my document unless, once again, they have the right fonts. If a character is commonly font in many fonts, you don't have to worry very much; if it is quite rarely included in fonts, you may have to provide a link to a font containing the right character.

So how does one know if a specific character is a) present on my system or b) commonly or rarely found in fonts? There are several free tools and techniques you can use (these are windows based tools: I'd be interested in hearing about linux or mac based ones).

One way of finding out if you have the character at all is to create a simple HTML document with the character and load it into a Mozilla-based browser (Netscape, Mozilla, K-meleon). These browsers have an algorithm they use to try and discover if any font on your system has characters required by your document but missing in the default font. It isn't perfect, but it is pretty good. If any font on your system has the character, these browsers are likely to find it.

What they won't tell you is how common the character is or which fonts have the character you are looking for. To find out which specific fonts have your character, I recommend a tool built by Conal Tuohy, "Find Glyph." This is a small, quick, and simple utility that allows you to find out which fonts have the Unicode point you are looking for. I used this to find out that I have two fonts with wynn, for example. Find Glyph is available for free download at http://www.nzetc.org/downloads/find-glyph.zip.

Another useful utility which seems to come and go on the web is Bjondi International's Character Agent. This won't tell you which fonts have a given character (I should be using the correct term, which is glyph, actually). But it does allow you to scroll through all the fonts on your system at a given unicode point regardless of whether the font in question has the character you are looking for (this is better than Character Map, a Windows utility, as Character Map only shows points in a font that have an actual glyph present). That was how I found out that Wynn is located in the Latin Extended B subrange. Character Agent is free, but I wasn't able to find a working link to it right now after a very quick search (Google throws up some links, but they seem to be outdated).

Does anybody know of other handy tools?

-dan